There’s normally a somewhat quite period in terms of voting intention after an election. There’s just been an actual vote, newspapers have blown all their polling budget during the campaign and even pollsters have to have a holiday. Sample quotas and weights all have to be rejigged as well (that applies even when polls have got an election correct – most polls’ quotas or weights include voting at the previous election, so 2015 targets all need replacing with 2017 targets).

We’ve had two Survation polls earlier this month, both showing Labour leads. Yesterday’s Sunday Times also had a new Panelbase poll, their first since the general election, and also showed Labour ahead. Topline figures there are CON 41%(-3), LAB 46%(+5), LDEM 6%(-2), changes are from the actual election result (or at least, the Great British vote share at the general election – the vast majority of opinion polls cover Great Britain only, not Great Britain and Northern Ireland).

It’s an interesting rhetorical question to ponder how much of the shift in public opinion since the election is because of the general election result (Theresa May’s figures have dropped now she is the PM who called a snap election and lost her majority, Jeremy Corbyn’s have shot up now he is a leader who deprived the Tories of a majority when he’d been so widely written off), and how much is the continuation of trends that were already there in the general election campaign? In other words, if the election had been a week later, would the trend towards Labour have continued and would they have been the largest party (or the Tories less able to form a viable government?). We’ll never know for sure.

605 Responses to “Panelbase/Sunday Times – CON 41%, LAB 46%, LDEM 6%”

1 2 3 13
  1. first?

  2. First?

  3. Definitely a rhetorical question, but the question of whether a shorter campaign (till May) would have been beneficial for the Tories is perhaps more pertinant for them.

    A deal with the DUP worth £1bn has just been announced. This seems, on the surface, a risky play for both the Conservatives and the DUP. For both there is the already much mentioned risk of association, but more acutely there is the message this send out about Conservative fiscal responsibility… And then subsequent discussions about devolution/local govt settlements…

    And for the DUP, is this deal worth nothing if a power-sharing agreement is not worked out soon?

  4. Not sure the public will be happy with the Tory/DUP deal which means £1.5 billion extra for Northern Ireland. If Scotland, Wales and indeed England don’t receive appropriate sums of additional money, then we might see further polls shift towards Labour.

    This also does not bode well for Brexit, as the Tories have had to dump policies, as part of the deal. No doubt compromises will have to made with Brexit and many who supported Brexit will not be happy.

  5. Wonderong why the deal took so long to finalise. There’s nothing obviously contentious, and £1bn or so is on the low end of expectations. Most notably, there doesn’t seem to be anything about how Brexit will affect the NI borders with ROI and GB.

    One has to wonder whether the real discussions were about things that are NOT included, and which the government will NOT be doing.

    There was talk of a legal challenge to the agreement. Perhaps the delay was in trying to find a way to disguise as much as possible the obvious conflict of interest implicit in the agreement.

    Thursday is now the interesting day – the deadline for restoring power sharing or ijmposing direct rule. If there is no restoration of power sharing, it seems impossible for the government’s action on Thursday to be seen as impartial, no matter what it is.

  6. “Wonderong why the deal took so long to finalise.”

    Reading between many of the lines left by commentators, I would counsel that this deal needed not just to be sold to the DUP, but also to the Tory backbenches.

    The country needs a functioning Government, and time will tell if we now have one.

  7. When will be the graphs and tables be updated with this new polling information?

  8. Tables aren’t out for the Panelbase poll yet – they normally don’t appear till later on the Monday at least when they do one for the Sunday Times. They should appear on here when they do:

    As already discussed what isn’t happening here is a ‘winner’s bonus’ and certainly not the start of a honeymoon period. Or if there is a winner’s bonus the public have awarded it to the runner-up.

    Whether this is due to post-election events or the momentum (or Momentum) of previous trends is probably not distinguishable, and may even be different aspects of the same things. For example May’s initial mishandling of the London fire is only a continuation of poor response to previous events. Not that it had much direct effect on VI I suspect (personal ratings may have been more affected), but it could have altered media perception in a similar way to the election campaign and result.


    Not sure if anyone already linked to it, but Ipsos Mori have released their “how they voted in 2017”
    Well worth a read

    Colin mentioned in a previous thread and it’s very interesting indeed, but it’s worth warning about its limitations. MORI’s figures, as they are the first to remind us, are based entirely on their own polling on the run up to the election. So it isn’t really How Britain Voted, but how some people said they might vote, spread over the seven weeks before.

    According to the commentary:

    the sample is based on 7,505 GB adults aged 18+ (of whom we have classified 5,255 as likely voters, using the same definitions as in previous election estimates), interviewed by telephone and online between 21 April – 7 June 2017. I reckon 4397 were from their published phone polls and combining phone and online polls is rare (though not unknown, Asdhcoft’s was also mixed mode) and may have hidden problems. The total sample is smaller than either Ashcroft or YouGov, both of which are of actual reported votes and it means that the figures from some of the subsamples will be based on a few hundred respondants if that.

    Also turnout is based on whether the likelihood to vote of those polled[1] which has then been calibrated to match the final actual results and turnout at a regional level, which should make them a more accurate guide to how different sub-groups voted. So that’s also a bit artificial. There doesn’t seem to be any input from the BBC Exit Poll, which MORI helped with.

    The great thing about this exercise though is that MORI has been carrying out this every election since 1979, without a lot of change in methodology. So when you’re comparing changes over time (as Richard did) it can be quite useful despite all the reservations. But there is a tendency for journalists to quote the MORI figures as if they were gospel and you do need to understand the limitations, which MORI do make clear.

    [1] Of course, as I said in a previous thread, if MORI had actually just done this on polling day rather than apply all sort of corrections, they would be up there sharing the smugfest with Survation.

    (Put up on previous thread as it died, but as it doesn’t relate to the topic of either it might as well go here too).

  10. Good Afternoon everyone from a sunny Bournemouth East.

    I think that the DUP-Tory pact may well bring stability to the Govt for some time ahead.

  11. The sudden discovery of the much-derided magic money tree by the Conservative Government will make for interesting politics in Scotland and Wales.

    In particular, there could be many difficulties for the Scottish Tory MPs. No doubt the SNP will demand extra money as well. What do the Scottish Tories do then?

  12. “In particular, there could be many difficulties for the Scottish Tory MPs. No doubt the SNP will demand extra money as well.”

    How curious. Do the SNP now want to vote with the Government as well?

  13. Robin

    My guess is that a deal to restore power sharing is pretty much sorted. That May have been why there was a delay to the Tory-DUP deal – to wait until a power sharing deal looked in the bag.

    On a lighter note, can we have a debate on what to call he deal. I find having to type Tory-DUP a pain on my IPAD

    How about “Trup”




    It sounds like [MORI] normalised it relative to the actual result, so it seems unlikley that they could have predicted the result without the correct weightings!

    I was referring to the final poll they released on polling day rather than any post-result adjustments. The point was how well the only two phone polls (MORI and the final Survation) had done.

    Amusingly MORI’s raw sample having exactly the same number of Con and Lab voters (439) but after MORI adjusted for “All 9/10 certain to vote + always/usually vote/depends + turnout overclaim adjustment + refused reallocated”, they ended up with:

    Con 44%, Lab 36%, LD 7%, UKIP 4%.

    However if you ignore all their adjustment except using 9-10 likely to vote (see column 8):

    you get:

    Con 42%, Lab 40%, LD 7%, UKIP 2%.

    pretty near the actual result:

    Con 44%, Lab 41%, LD 8%, UKIP 2%.

    and to Survation’s

    Con 41%, Lab 40%, LD 8%, UKIP 2%.

  15. Does any one know if Panel base (and other companies with a couple of honourable exceptions) ) have changed their methodology after getting it badly wrong in the General Election.
    If the haven’t changed their methodology could the Labour lead be substantially higher?

    Will there be a further polling enquiry to try and establish the reasons or is it pretty clear cut that that the reason was some sections of the electorate voted in larger numbers than previosly.

  16. @BT

    “How curious. Do the SNP now want to vote with the Government as well?”

    How curious. Do you think the SNP will say, “Righty-ho. Fair enough. Give Northern Ireleand an extra £1 billion. That’s ok with us. We won’t ask for any more.”?

  17. @Chrislane1945
    Are you trying to do a reverse outcome gambit by predicting stability for the DUP-Conservative deal? ie if you predict it, it won’t happen.
    Seems to me both sets of coats are on ‘Shugley pegs’!

  18. Who do we think will be picking up the bill for the £1 billion that the Govt have just promised to keep themselves in number 10.

    The NHS?
    Social Care?

    Or will the govt ask the taxpayer to stump up more tax?

    There seems something a little wrong about paying out cash to keep your majority,

    Cash for votes?

  19. Norbold

    I think it can easily be shown how much the Barnett formula ‘generously’ (in our language, unfairly) rewards Scotland already.

    I also think the SNP would far rather have another grievance to beat Westmonster with for the next 4 years than the £1b.

    Incidentally, is that £1b for NI a one-off or an addition to the annual grant?

  20. @TonyBTG

    I see nothing that would persuade Sinn Fein to reverse their position concerning Foster’s alleged complicity in the Ash for Cash scandal.

  21. @ Chris Lane

    “I think that the DUP-Tory pact may well bring stability to the Govt for some time ahead.”

    I don’t share your optimism. The DUP has traditionally received support from the UDA loyalists, who have implicated in the killings of about 400 citizens of Northern Ireland, mostly Catholics. That’s like Labour forming a Coalition with Sinn Fein.

    I don’t see this ending well….

  22. Tony BTG

    What sort of deal would you have expected, if it were not, ultimately, to be ‘cash’ for something or other?

    Just curious what non-monetary benefits you think the DUP / NI would be so keen for.

    Also, ‘Cash for Votes’ – either you didn’t realise the £1bn wasn’t payment to DUP MPs for their votes, it was for the people of NI, or I’ve missed something very big and cr*p is about to hit the fan!

  23. @ NeilJ

    “Does any one know if Panel base (and other companies with a couple of honourable exceptions) ) have changed their methodology after getting it badly wrong in the General Election.
    If the haven’t changed their methodology could the Labour lead be substantially higher?”

    Indeed…I don’t have much confidence in Panelbase right now.

    That said, Survation have a similar five-point lead for Labour, and Survation were arguably one of the more accurate pollsters for the GE.

  24. “That’s like Labour forming a Coalition with Sinn Fein.”

    If you’re in Irish catholic it might be seen that way.

    It’s certainly not how Jo Public in GB who well remember the Troubles, see it. They don’t forget who the IRA are / were, but if you asked them who the UDA were without context, I expect most would guess a pop group or local health clinic.

  25. I suppose there is a very real danger to the Tories that they could be setting fire to their hard won ‘economic competence’ polling lead with the DUP deal.

    Am I the only one that is reminded a bit of the ERM when the govt. spent a fortune propping up the pound?

    Having campaigned on ‘no magic money tree’ / ‘no alternative to austerity’, ‘public sector wage freezes are necesary’ etc it looks particularly bad that they can suddenly find this extra money when needed.

    It looks a bit like they just found the magic money tree but it only grows in Northern Ireland.

    I can’t imagine that narrative having a positive effect for Scottish Tories either.

    Will be interesting to see the next opinion polls after the dust settles from this deal.

  26. tonybtg

    “On a lighter note, can we have a debate on what to call he deal. I find having to type Tory-DUP a pain on my IPAD”

    Coalition of Chaos? or chicken for short?

  27. @BT

    “What sort of deal would you have expected, if it were not, ultimately, to be ‘cash’ for something or other?”

    Hence the conundrum that both sides find themselves in… All that the Tories are doing is buying themselves time, and I don’t know what game the DUP are playing, but power-sharing seems impossible with Arlene Foster as leader, at least without admitting some culpability in the “Cash for Ash” RHI debacle.

    The Tories seem to be continually digging their hole deeper and deeper, calling it quits sooner would possibly be to their advantage… As I’ve said before, defections may be on the card, especially if the Queens speech is passed (a failure may result in an even bigger constitutional mess, so lets deal with that when/if it comes…)

  28. Neil –

    Not yet. Obviously everyone (apart from MORI, who don’t use political weighting) will update their political weighting to 2017 rather than 2015. The question is what happens to turnout models that were previously turning small Tory leads into great big ones. My guess is that companies will drop them (as Roger says in his comment above, lots of polls would have been pretty accurate without them!), at least for the time being.

    Longer term we shall see what happens. Dropping those turnout models is not necessarily a solution, without changes polls would still have been wrong in 2015, the problem of samples that are too engaged hasn’t necessarily gone away… it just wasn’t a problem in 2017 (or perhaps it produced a skew in another direction, so using a turnout model that bumped up Tory voting groups exacerbated the problem, rather than making it better). Essentially, if turnout remains high then methods similar to those from 2015 will probably produce correct results; if turnout drops again (or the pattern of turnout changes) then they may not. The problem many polls made this time round was assuming turnout in 2017 would resemble turnout in 2015. Any solution that involves assuming turnout in 2022 (or whenever) will be the same as in 2017 carries similar risks.

    Though just to upset all the assumptions I’ve just made about turnout being the problem… the YouGov MRP model used turnout among different demographics in 2015 as factor in the model… so while it was obviously a major cause of the error in ComRes and ICM’s polls, don’t assume anything using past turnout as a guide is going to be wrong!

  29. @BT

    “I think it can easily be shown how much the Barnett formula ‘generously’ (in our language, unfairly) rewards Scotland already.”

    And you think the SNP will go along with that do you?

  30. I quite like Con-D as an acronym for the deal

  31. @theexterminatingdalek

    DUPCon? CoDUP? Perhaps just “Zelda”??

    I cannot believe anyone (Chris Lane) seriously believes that this brings stability… In what way does this bring stability? The country is swinging away from the Conservatives at an almost unprecedented rate, and NI has been in a political crisis since Martin McG resigned, despite the national press barely reporting on it during the election.

    The irony of the “Strong & Stable” mantra would be hilarious were it not so serious. On a (tenuous) polling note, will this national risk-taking for personal/party political purposes not serve to only undermine th DUP and the Conservatives further? That is almost the only consolation, that the stability is so poor, that it cannot hang on, perhaps even until the summer recess.

  32. or maybe Condup

  33. “And you think the SNP will go along with that do you?”

    Of course not. But it’s demonstrable enough not to lose more votes to the SNP in the long term.

  34. JIM JAM (FPT)

    Personally I would be surprised at UKIP 2015 going 60% Can and 16% Lab in the end, it maybe that early responses skew the average. Nearer 50/20 would be my take.

    MORI’s figure is pretty close to Ashcroft’s ones from a sample of 1646 UKIP voters in 2015:

    and YouGov’s figures from 4646 UKIP voters are not much different:

    Con 57% [60] {64}

    Lab 17% [16] {16}

    Lib Dem 3% [1] {2}

    UKIP 19% [18] {14}

    Green[1] 2% [All other 5] {1}

    Other 2% {2}

    Ashcroft []=MORI {}=YouGov

    Though both Ashcroft and MORI don’t examine how many 2015 UKIP voters didn’t vote this time. YouGov claim 30% of the total, though I have severe doubts over the related weighting. But a higher figure than for other Parties is not unlikely.

    What these polls do make clear is that much of the talk of Labour benefiting to any large extent from the UKIP collapse – getting a half, third, or even a quarter of the defectors – seems not to have happened.

    [1] One of the delights of Ashcroft’s figures is how they illustrate how successfully Labour squeezed the Greens. In 2017 more voters moved to them from Con than Lab and from UKIP than Lib Dem.

  35. BT

    So that will be the Scottish Tory MPs argument will it? We don’t need more money. Good luck to Northern Ireland. Do you think that will go down well in SCotland?

    Incidentally, Plaid Cymru’s Westminster leader Liz Saville Roberts said: “Any commitments for Northern Ireland should be matched for Wales.

    “If reports that the DUP has secured a £1bn increase in public spending in Northern Ireland are realised, Wales’ population share would be around £1.7bn – a substantial boost to the Welsh economy that must be delivered.”

  36. @BT

    And Barnett rewards NI even more generously. It has the highest public expenditure per head in the UK and the worst fiscal balance before this deal.

  37. Norbold

    You’re being silly. We don’t need on this thread to predict the wording each side of the debate will use, it’s obvious they will have the political equivalent of a punch-up about it. What your further point is I don’t know.

  38. @GrahamBC

    I suspect the headline writers will be working away feverishly now and whatever they come up with will frame the deal.

    “Cons are DUPed” seems an obvious one though.

  39. @BTsays

    A different way of saying what Norbold istrying to get across is that all non-Tory Scottish MPs will now be VERY loud in asking Scottish Tories why they haven’t negotiated for an equivalent increase in Scottosh funding. To which the answer will be what? “Yes, we support giving more money to NI but not to Scotland”?


    Or “DUP are Conned” depending on your POV.

  41. Robin

    Yes, I understood that thank you.

  42. @PETE B

    Indeed. It won’t get much attention here in the UK but I wonder what the impact will be on the DUP in NI in the long term. They might have their own backlash to deal with.

  43. @ RogerMexico

    What puzzles me about this UKIP apportionment is how it seems to have been very uneven in those semi “northern heartlands”,

    After those Sunderland results came in I did wonder whether the exit poll was going to be right. Yet Sunderland was not repeated in, say, Halifax, where a 13%/5,600 UKIP vote in 2015 didn’t translate into a Tory gain when they just needed a few hundred extra votes over and above any extra Lab votes.

    It was the same pattern throughout and it almost felt to me like the smaller the Lab majority, the less or opposite the swing was.

    My only theory is that some squeezing of the UKIP vote had already been done in 2015 by the Tories in seats where it mattered so they had less potential Tory UKIP voters to eat into. Whereas the bigger the Lab majority the less UKIP squeezing had gone on in 2015 and Tory voters just naturally drifted back.

  44. And to answer my own earlier question, it looks like the £1bn to NI is a one-off for the next two years not an addition to the usual grant, if you look at what it’s for (one off costs relating to Health, infrastructure and deprivation in communities).

    Which is a relief IMHO, as that means it’s basically peanuts compared to anybody’s manifesto commitments for ongoing new costs (or cost savings) on anything, year on year.

  45. That should have read…’in the rest of the UK’ before everyone from NI jumps on me. :-)

  46. Robin – it was pretty obvious what Norbold meant, applies to Wales as well of course and then ‘England’ may say how come all except us?

    I think Chris is right and I expect the DUP deal to last at last until November 2018 and probably spring 2019 for am interim deal to be passed by the HOC.

    The DUP not supporting said interim deal may mean Nov 2018.

  47. @Anthony Wells

    “the problem of samples that are too engaged hasn’t necessarily gone away…”

    And this is probably the crux of the issue. A good turnout model is important, but good sampling should make that unnecessary. Understanding of the size of the “iceberg” is important. Seeing those that are really active can be easy, but 2017 seemed to under-estimate the extent to which the new Labour membership reflected a longer-term trend of traditionally less-engaged citizens affiliating themselves strongly to the Labour party.

    All the talk in the media was about how Labour supporters over-hyped this support, and yet with hindsight the unprecedented growth in the Labour party should have served as a clue as to the size of the Labour “iceberg” below the surface! Whilst the habitual Tory voter, or slightly concerned conservative who may have previously voted for another party is obviously still there given the share of the Tory vote, this was thought to contribute to a bigger “iceberg” than ultimately transpired.

    When things change fast, I tentatively suggest that accurate polling without vast resources is probably impossible. All the results demonstrated was the biases of the polling companies, presumably just well-intentioned mis-judgements as I seriously doubt that polling companies would want to actively overstate the VI share of one party over another. Would they? Perhaps unintentially I suppose?

    I know that this may seem malicious to some, but I am just being devils advocate: Can Lord Ashcroft ever be judged impartial without publishing his methodology in full, including the sampling strategy, as he is a politician himself?

  48. ” We’ll never know for sure.”

    But if Anthonly provided a graph showing the rise and rise of Corbyn as we had seen it on YouGov and then the Survation and Paneblase figures for its post-election continuation, we’;ld have a pretty good idea that it was a trend.
    And in that case – in “innovation adoption” theory (Everett Rogers) – that increased awareness among those with access to knowledge of the recommendation package – for whch, read manifesto – meant that late adopters – DKs and Didn’t Votes – have been increasingly coming on board.

  49. BT

    “You’re being silly. We don’t need on this thread to predict the wording each side of the debate will use, it’s obvious they will have the political equivalent of a punch-up about it. What your further point is I don’t know.”

    I haven’t got a further point. However, the point I originally made, which you seem to like dodging around, is how will this play with the 13 Tory Scottish MPs. It was you who started stating the obvious about the SNP etc.

  50. SF haven’t reacted yet, I wonder how they are going to play it. They will probably feel a lot of pressure to do a deal on the assembly so that they can ensure that their community gets some of the extra money. Arlene foster is off the hook for the cash for ash scandal, there is no way SF can block her from being first Minister and having control over the inquiry about that scandal. At least we know who won the election now.

1 2 3 13